What we ought to be thinking about is not the individual bacterium as the unit of evolution, but these bits of DNA as the units of evolution. It's an idea that will be familiar to you, of course. And, let me give you one example of it, it's not my idea at all, it's suggested by two Canadian microbiologists, [JG] Lawrence and [JR] Roth. I think it's a fascinating notion. There are, in bacteria, groups of genes called operons. The classic one is the so-called lac operon in E. coli, it's the first one to be discovered. It's a group of genes, which if you've got them, they're all end-to-end on the chromosome, and if you switch them on, they enable the cell to digest and cope with lactose. And lactose is a sugar which E. coli doesn't often meet, but if it does, it can say, gosh, I must do something, and it can switch on these genes and start digesting lactose. And it was the discovery of this group of genes, and in particular the discovery of how the presence of the lactose switched the gene on, which was done by the group at the Pasteur, particularly Jacques Monod and François Jacob, which really was the beginning of our understanding of how genes are regulated. I mean, for me, after the Watson and Crick paper, the Jacob - Monod work was the most important thing that's happened in biology in my lifetime. But nevertheless, there's a puzzle. You've got this group of genes, and there are many operons in E. coli and every bacterium has these operons, groups of link genes with a switch or switches at the beginning for switching them on, when you need them. Why are they linked? You see, the thing would work perfectly well if the 10 genes, or whatever it is, of an operon, were spotted all around the chromosome, each with its little switch, and when you want them, you switch them all on, and they can all have the same switch and switch on. They don't have to be together. Monod would have replied, 'Well, it's easier to regulate them as a single switch,' and that's a perfectly plausible explanation. Maybe it's just they are together because it's convenient to switch them all on at once. What Roth and Lawrence point out is that actually being together has another advantage, it means you can be transferred from one cell to another together, and they want to suggest that these operons are little selfish... not genes, but selfish genetic elements. They've evolved under selection acting on them, because as a group, they can transfer. Now, this is interesting, because if it's true, there are certain implications. You wouldn't expect to find genes that you need all the time to be in operons, because you'd have them, you wouldn't need to acquire them from somewhere, they wouldn't need to move around. The genes you'd expect to find in operons would be the genes which are only occasionally needed, only in special environments. And then, if you can work it out, there is selection in favour of tight linkage. And what Roth and Lawrence show is that, by and large, that prediction is born out, that genes like the lac operon, for lactose, which are... for an environment which is only occasionally needed, are in operons linked together, genes for DNA replication are not. You need to replicate your DNA all the time, and they're spotted all around the chromosome, switched on together, but they're not linked. And there are other reasons for thinking that maybe this sort of gene-centred view of bacterial evolution may be the right way of looking at it. The one I'm particularly interested in, and I want to work on it with a young friend in Paris, is the possibility that the genes responsible for transformation, this process of sex that I was describing earlier on, may be the same, that sexual transfer in the bacteria may be being carried out by groups of genes which are really doing it for their own good, because they want to jump, you know. They're bored with sitting in the same cell, they'd like to replicate, like getting into new cells. I don't know whether this'll turn out to be right.